贝叶斯共同成分模型拓展及其在高血压风险评估的应用
|
摘要
尝试引入贝叶斯样条对贝叶斯共同成分模型进行拓展,并基于1991-2011年《中国健康与营养调查》的实际数据探究高血压风险的空间变异与时间变异。研究发现:高血压风险(尤其是男性)大致呈现出北方高于南方的空间格局,且这一空间分布模式具有相对稳定性,这可能与北方各省份相似且稳定的生活方式(饮酒、膳食高盐等)有关。拓展后的SCM模型更充分地挖掘样本信息,揭示出高血压风险的空间格局及其随时间演变的规律,克服了空间分析时样本量相对不足的局限。
This paper tried to extend the existing Shared Component Model(SCM)by introducing the Bayesian spline and employed the extended model on the basis of 1991-2011 China Health and Nutrition Survey(CHNS).Results showed that the risk of hypertension(especially for males)generally showed a spatial pattern that is higher in the Northern China than that in the Southern China.Also,this pattern was relatively stable during the study period.This pattern may be related to the similar and stable lifestyle in the Northern areas,such as drinking alcohol,high salt diet,etc.The extended SCM is able to explore the sample information fully and reveal the spatial pattern and trend over time with respect of the risk of hypertension,thus overcome the limitation of limited sample size of spatial analysis.
This paper tried to extend the existing Shared Component Model(SCM)by introducing the Bayesian spline and employed the extended model on the basis of 1991-2011 China Health and Nutrition Survey(CHNS).Results showed that the risk of hypertension(especially for males)generally showed a spatial pattern that is higher in the Northern China than that in the Southern China.Also,this pattern was relatively stable during the study period.This pattern may be related to the similar and stable lifestyle in the Northern areas,such as drinking alcohol,high salt diet,etc.The extended SCM is able to explore the sample information fully and reveal the spatial pattern and trend over time with respect of the risk of hypertension,thus overcome the limitation of limited sample size of spatial analysis.
引文
[1]刘力生.中国高血压防治指南2010[J].中华高血压杂志,2011,(8):701-743.
[21]Who W H O.A global brief on Hypertension-Silent killer[R].Global Public Health Crisis,2013.
[3]罗雷,栾荣生,袁萍.中国居民高血压病主要危险因素的Meta分析[J].中华流行病学杂志,2003,(1):56-59.
[4]李新建,徐继英,姚海宏.上海市居民高血压危险因素病例对照研究[J].中国慢性病预防与控制,2010,(4):375-376.
[5]Niskanen L,Laaksonen D E,Nyyssonen K,et al.Inflammation,abdominal obesity,and smoking as predictors of hypertension[J].Hypertension,2004,44(6):859-865.
[6]王丽娜,曹丽,张敬一,等.河北省成年居民高血压患病状况及相关危险因素分析[J].中国慢性病预防与控制,2008,16(2):125-129.
[7]肖静,杨梅,宗蕾,等.高血压健康危险因素的logistic回归和对数线性模型分析[J].中国慢性病预防与控制,2012,20(4):372-374.
[8]刘红艳,彭辉,刘奥博,等.中国人原发性高血压危险因素Meta分析[J].中国心血管杂志,2009,14(3):205-209.
[9]Kagan A,Faibcl H,Bcn-Aric G,et al.Gonder differences in ambulatory blood pressure monitoring profile in obese,overweight and normal subjects[J].Journal of Human Hypertension,2007,21(2):128—134.
[10]Pechere-Bertschi A,Burnier M.Female sex hormones,salt,and blood pressure regulation[J].American Journal of Hypertension,2004,17(10):994-1001.
[11]Sesso H D,Cook N R,Buring J E,et al.Alcohol consumption and the risk of hypertension in women and men[J].Hypertension,2008,51(4):1080-1087.
[12]Cappuccio F P,Stranges S,Kandala N,et al.Gender-specific associations of short sleep duration with prevalent and incident hypertension the Whitehall II study[J].Hypertension,2007,50(4):693-700.
[13]Knorr Held L,Best N G.A shared component model for detecting joint and selective clustering of two diseases[J].Journal of the Royal Statistical Society:Series A(Statistics in Society),2001,164(1):73-85.
[14]Dabney A R,Wakefield J C.Issues in the mapping of two diseases[J].Statistical methods in medical research,2005,14(1):83-112.
[15]Held L,Natário I,Fenton S E,et al.Towards joint disease mapping[J].Statistical methods in medical research,2005,14(1):61-82.
[16]Richardson S,Abelian J J,Best N.Bayesian spatio-temporal analysis of joint patterns of male and female lung cancer risks in Yorkshire(UK)[J].Statistical Methods in Medical Research,2006,15(4):385-407.
[17]Tzala E,Best N.Bayesian latent variable modelling of multivariate spatio-temporal variation in cancer mortality[J].Statistical Methods in Medical Research,2008,17(1):97-118.
[18]Baker J,White N,Mengersen K,Rolfe M,Morgan G G.Joint modelling of potentially avoidable hospitalisation for five diseases accounting for spatiotemporal effects:A case study in New South Wales,Australia[J/ON].PLoS ONE,2017,12(8):e0183653.DOI:10.1371/journal.pone.0183653.
[19]Mahaki B,Mehrabi Y,Kavousi A,et al.A spatio-temporal multivariate shared component model with an application to Iran cancer data[R].arXiv preprint arXiv:1707.06075.2017.
[20]王晓荷,田茂再.基于分层贝叶斯时空模型的雾霾天气过程计数分析[J].数理统计与管理,2017,36(6):970-982.
[21]苏珊娜.霍夫曼(Susanna Hoffman)(著),陈梅(译).人类学视野中的灾难和文化要素的探讨[J].Journal of Ethnology,2015,(4):29-35.DOI:10.3969/j.issn.1674-9391.2015.04.005.
[22]Spiegelhalt.er D J,Best N G,Carlin B P,et al.Bayesian measures of model complexity and fit[J].Journal of the Royal Statistical Society:Series B(Statistical Methodology),2002,64(4):583-639.
[23]MacNab Y C.Spline smoothing in Bayesian disease mapping[J].Environmetrics,2007,18(7):727-744.
[24]Ruppert D.Selecting the number of knots for penalized splines[J].Journal of Computational and Graphical Statistics,2012,11(4):735-757.
[25]MacNab Y C,Gustafson P.Regression B-spline smoothing in Bayesian disease mapping:With an application to patient safety surveillance[J].Statistics in medicine,2007,26(24):4455-4474.
[26]Ancelet S,Abellan J J,Del Rio Vilas V J,et al.Bayesian shared spatial-component models to combine and borrow strength across sparse disease surveillance sources[J].Biometrical Journal,2012,54(3):385-404.
[27]Li Xu,Dejian Lai,Ya Fang.Spatial analysis of gender variation in the prevalence of hypertension among the middle-aged and elderly population in Zhejiang Province,China[J].BMC Public Health,2016,16:447.DOI:10.1186/s12889-016-3121-y.
[21]Who W H O.A global brief on Hypertension-Silent killer[R].Global Public Health Crisis,2013.
[3]罗雷,栾荣生,袁萍.中国居民高血压病主要危险因素的Meta分析[J].中华流行病学杂志,2003,(1):56-59.
[4]李新建,徐继英,姚海宏.上海市居民高血压危险因素病例对照研究[J].中国慢性病预防与控制,2010,(4):375-376.
[5]Niskanen L,Laaksonen D E,Nyyssonen K,et al.Inflammation,abdominal obesity,and smoking as predictors of hypertension[J].Hypertension,2004,44(6):859-865.
[6]王丽娜,曹丽,张敬一,等.河北省成年居民高血压患病状况及相关危险因素分析[J].中国慢性病预防与控制,2008,16(2):125-129.
[7]肖静,杨梅,宗蕾,等.高血压健康危险因素的logistic回归和对数线性模型分析[J].中国慢性病预防与控制,2012,20(4):372-374.
[8]刘红艳,彭辉,刘奥博,等.中国人原发性高血压危险因素Meta分析[J].中国心血管杂志,2009,14(3):205-209.
[9]Kagan A,Faibcl H,Bcn-Aric G,et al.Gonder differences in ambulatory blood pressure monitoring profile in obese,overweight and normal subjects[J].Journal of Human Hypertension,2007,21(2):128—134.
[10]Pechere-Bertschi A,Burnier M.Female sex hormones,salt,and blood pressure regulation[J].American Journal of Hypertension,2004,17(10):994-1001.
[11]Sesso H D,Cook N R,Buring J E,et al.Alcohol consumption and the risk of hypertension in women and men[J].Hypertension,2008,51(4):1080-1087.
[12]Cappuccio F P,Stranges S,Kandala N,et al.Gender-specific associations of short sleep duration with prevalent and incident hypertension the Whitehall II study[J].Hypertension,2007,50(4):693-700.
[13]Knorr Held L,Best N G.A shared component model for detecting joint and selective clustering of two diseases[J].Journal of the Royal Statistical Society:Series A(Statistics in Society),2001,164(1):73-85.
[14]Dabney A R,Wakefield J C.Issues in the mapping of two diseases[J].Statistical methods in medical research,2005,14(1):83-112.
[15]Held L,Natário I,Fenton S E,et al.Towards joint disease mapping[J].Statistical methods in medical research,2005,14(1):61-82.
[16]Richardson S,Abelian J J,Best N.Bayesian spatio-temporal analysis of joint patterns of male and female lung cancer risks in Yorkshire(UK)[J].Statistical Methods in Medical Research,2006,15(4):385-407.
[17]Tzala E,Best N.Bayesian latent variable modelling of multivariate spatio-temporal variation in cancer mortality[J].Statistical Methods in Medical Research,2008,17(1):97-118.
[18]Baker J,White N,Mengersen K,Rolfe M,Morgan G G.Joint modelling of potentially avoidable hospitalisation for five diseases accounting for spatiotemporal effects:A case study in New South Wales,Australia[J/ON].PLoS ONE,2017,12(8):e0183653.DOI:10.1371/journal.pone.0183653.
[19]Mahaki B,Mehrabi Y,Kavousi A,et al.A spatio-temporal multivariate shared component model with an application to Iran cancer data[R].arXiv preprint arXiv:1707.06075.2017.
[20]王晓荷,田茂再.基于分层贝叶斯时空模型的雾霾天气过程计数分析[J].数理统计与管理,2017,36(6):970-982.
[21]苏珊娜.霍夫曼(Susanna Hoffman)(著),陈梅(译).人类学视野中的灾难和文化要素的探讨[J].Journal of Ethnology,2015,(4):29-35.DOI:10.3969/j.issn.1674-9391.2015.04.005.
[22]Spiegelhalt.er D J,Best N G,Carlin B P,et al.Bayesian measures of model complexity and fit[J].Journal of the Royal Statistical Society:Series B(Statistical Methodology),2002,64(4):583-639.
[23]MacNab Y C.Spline smoothing in Bayesian disease mapping[J].Environmetrics,2007,18(7):727-744.
[24]Ruppert D.Selecting the number of knots for penalized splines[J].Journal of Computational and Graphical Statistics,2012,11(4):735-757.
[25]MacNab Y C,Gustafson P.Regression B-spline smoothing in Bayesian disease mapping:With an application to patient safety surveillance[J].Statistics in medicine,2007,26(24):4455-4474.
[26]Ancelet S,Abellan J J,Del Rio Vilas V J,et al.Bayesian shared spatial-component models to combine and borrow strength across sparse disease surveillance sources[J].Biometrical Journal,2012,54(3):385-404.
[27]Li Xu,Dejian Lai,Ya Fang.Spatial analysis of gender variation in the prevalence of hypertension among the middle-aged and elderly population in Zhejiang Province,China[J].BMC Public Health,2016,16:447.DOI:10.1186/s12889-016-3121-y.